Supporting Information
Table S1 : Dataset of soluble sugar (glucose, fructose and sucrose), starch and total non-structural carbohydrates.
Fig. S1 Graphical validation of the assumptions of linear model describing seasonal dynamic of soluble sugars (homoscedasticity, normality of the error distribution, statistical independence of the errors, absence of influent points).
Fig. S2 Graphical validation of the assumptions of linear model describing seasonal starch dynamic (homoscedasticity, normality of the error distribution, statistical independence of the errors, absence of influent points).
Fig. S3 Graphical validation of the assumptions of linear model describing the seasonal total carbohydrates dynamic (homoscedasticity, normality of the error distribution, statistical independence of the errors, absence of influent points).
Fig. S4 Multiple comparison of total sugars, starch, and soluble sugars performed by the Student-Newman-Keuls Method. On the axis the day of the year (DOY) and the year of the samplings were reported (DOY_Year). Point represent significant difference (P-value < 0.05).